ShowHN: Qwiki for your web accounts (GitHub, HN, Netflix, etc) | Hacker News Reader

ShowHN: Qwiki for your web accounts (GitHub, HN, Netflix, etc)

33 pointsdavidcann21 comments

Remember that Qwiki demo at TC Disrupt last month? I figured something like that might be useful, but for your morning routine of checking your various web accounts.

http://verbaljuicer.com

It's currently a Safari extension and uses the very good OS X text-to-speech engine.

There's a JavaScript API to add scripts for any other website you want to track. If you write a script for a new site, email it to me at david at davidcann.com and I'll add it to the gallery.

Stats: I made most of it in about 3 days, then polished it for another 3 days.

What do you think? Should I port it to Chrome?

View on HN

I'd love to use it, but I'm on Windows 7 with Chrome.

If you port it I'd be willing to pay $ for it, maybe have a free version with those default sites but a paid that allows more sites to be integrated and allow users to submit their integrated sites?

Well anyway, nice idea, shame I can't use it :-(

Yea, I'd like to port it to windows and chrome, but this was just my MVP.

Do you know if the Windows built-in text-to-speech is any good?

Wow! I am amazed by this.

Some out-loud thinking; I wonder if you can apply this to news delivery in full. What I mean is, If I could submit a URL of the NYT business column and request that every morning, your app email me the mp3 file covering the whole column or just one article.

Essentially I can be listening to my morning paper on the go rather then reading it or needing a browser.

Very cool app! Nicely done

Thanks! That's a great idea and they actually already produce an audio podcast of the NYT and WSJ each morning on Audible. I could definitely see automating it for other news sites and blogs, etc. In fact, that can already be added using the javascript API!

Its great to see others like Qwiki and verbaljuicer getting and or interested in the alarm clock content space. Our alarm service(sleep.fm) speaks weather reports and soon(http://sleep.fm/coming-soon) wakes and tells you if your airline flight is delayed or on-time.

Though we only speak and repeat up to 8 seconds of alarm info. Were interested to see in time how this upcoming industry plays out. Does the general population want to hear just a buzzing noise, 5 to 8 seconds of personalized alarm info or hear an alarm voice reading 10 minutes or more of alarm personalized information? What can people grasp upon waking up?

Myself, I set three alarms and finally by the 2nd or 3rd alarm im able to grasp the weather condition and what the temperature will be around.

Id love to try this when you have a windows version available. Good luck!

Sleep.fm looks pretty cool. I can see that working great as an iPhone app alarm clock.

If you don't mind me asking, what are you using for your text-to-speech engine? Google's? A licensed web service? or something on your own server?

thnx.

We're using something we created on our server to stream to the site and our iphone app. Using Google's free 411 type engine would be awesome though; yet not sure if you can mix music into the background of the voice with theirs.

This is awesome, please port it to Chrome.

Also, any chance you can release the source for the page plugins you've already written? A couple aren't working for me (for example I need the weather to search google.co.uk not google.com) and it would be nice to be able to just make a quick edit rather than starting anew.

For anyone else looking for the plugin sources, they are embedded in the Verbal Juicer homepage.

With Firebug/Webkit dev tools, navigate to: body > #appsCode and the plugins and you can copy-paste into verbal juicer's create window.

For UK users, changing the weather code from google.com to google.co.uk seems to be all you need to do.

Actually, it's even easier than that! You can edit every plug-in right within the interface. Click the Edit button, then click "Go" on whichever plug-in script you want to edit. That page will load and the Go button changes to an Edit button. Click that and the code editor interface pops up. Change the URL and click Save!

I should redesign that Edit/Go button interface to make it more clear. A couple people have thought they can't edit the plug-in scripts.

I stumbled across that later, much easier :).

I seem to be getting a weird bug where after a while the code editor window pops up with 'undefined' in it and I can't do anything further. Uninstalling and reinstalling the extension seems to fix it, but I lose my plugins. Sorry I can't provide much more detail than that, if it happens again I'll see if I can figure anything else out.

An even more awesome feature request than porting it to Chrome for me would be letting me use JQuery in the plugins.

p.s. If this functionality was wrapped up into a slightly more stable standalone app, with really basic alarm clock functionality, I would pay for it in a heartbeat.

Eek, sorry about that bug. I'll look into it. Admittedly, I haven't done extensive or automated testing yet.

I agree JQuery would be a good addition and not difficult to add.

A standalone app would be nice too... that would make it easier to preload content and minimize loading delays between pages! A standalone app would also give me access to AppleScript, so it could integrate with Mail, iCal, etc.

I can seem to repeat the bug if I: - Open Safari - Install Verbal Juicer - Edit one of the plugins - Quit Safari whilst the code-editor is still open.

When I reopen Safari and try and edit an extension, I just get "undefined".

This great, and I am already thinking of plugins to add to it. The speed it takes to switch from one section to another could easily be increased by loading the list of sites on tabs in the background as it is reading off the data.

And, yes please port it to Chrome.

That's a great idea... I'll have to look into what control extensions have over tabs. I agree that the page load time is currently a weakness.

Just watched the demo video. Cool stuff.

How does it handle authentication? Based on an existing cookie?

Thanks! Yea, it's a browser extension, so it injects javascript into the page.

Your data is never transmitted to a server... it's all done locally in javascript and the text-to-speech is done locally via the plug-in.

Looking at your HTML Code:

- Don't pollute the global namespace

- Use Event binding instead of attribute events. Avoid inline JavaScript

- Delete comments before deploying (and compress HTML, JS, CSS)

Nice and simple design. Would like to see how this works in Chrome.

Thanks for the feedback, csomar. I will definitely be cleaning up and streamlining the code when I make it cross-browser and cross-platform.

Clickable: http://verbaljuicer.com/

This looks pretty damn awesome. Please port it to Chrome!

Reply on news.ycombinator.com